Table 1.

A comparison of PRC and XPRC at different precision levels determined by the PRC algorithm on the Genomics dataset. The cumulative article count is the number of articles with PRC precision below a given precision level. For example, there are 158 articles that result in PRC precisions lower than 0.2. PRC has precision 0.0 on all the 58 articles in the 0.0 group, so its macro-average precision and standard error are also 0. XPRC has better performance on these articles. p-value shows the significance of the difference between PRC and XPRC at different precision levels.

Precision levels		0.0	0.2	0.4	0.6	0.8	1.0
Cumulative article counts		58	158	314	603	1215	4234
PRC	macro-average precision	0	0.127	0.262	0.424	0.613	0.889
	standard error	0	0.008	0.009	0.008	0.007	0.003
XPRC	macro-average precision	0.338	0.443	0.523	0.614	0.705	0.864
	standard error	0.048	0.028	0.019	0.013	0.009	0.004
p-value		3e-09	3e-21	3e-30	2e-31	2e-16	4e-07